0084-6570/99/1015-0509$12.00 509

Annu. Rev. Anthropol. 1999. 28:509–29Copyright © 1999 by Annual Reviews. All rights reserved

THE HUMAN ADAPTATION FOR CULTURE

Michael TomaselloMax Planck Institute for Evolutionary Anthropology, Inselstrasse 22, D-04103 Leipzig,

Germany; e-mail: tomas@eva.mpg.de

Key Words: culture, cognition, ontogeny, primates, language

n Abstract Human beings are biologically adapted for culture in ways thatother primates are not, as evidenced most clearly by the fact that only humancultural traditions accumulate modifications over historical time (the ratchet ef-fect). The key adaptation is one that enables individuals to understand other in-dividuals as intentional agents like the self. This species-unique form of socialcognition emerges in human ontogeny at approximately 1 year of age, as infantsbegin to engage with other persons in various kinds of joint attentional activitiesinvolving gaze following, social referencing, and gestural communication.Young children’s joint attentional skills then engender some uniquely powerfulforms of cultural learning, enabling the acquisition of language, discourseskills, tool-use practices, and other conventional activities. These novel formsof cultural learning allow human beings to, in effect, pool their cognitive re-sources both contemporaneously and over historical time in ways that areunique in the animal kingdom.

CONTENTSIntroduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 510

Primate and Human Cognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511

The Ontogeny of Human Cultural Learning . . . . . . . . . . . . . . . . . . . . . . . . . 513

Joint Attention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513

Imitative Learning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514

Linguistic Symbols and Cognitive Representation. . . . . . . . . . . . . . . . . . . . . . . . . 516

Nonhuman Primate Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518

Japanese Macaque Potato Washing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519

Chimpanzee Tool Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519

Chimpanzee Gestural Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521

Nonhuman Primate Social Learning and Social Cognition . . . . . . . . . . . . . . . . . 523

Enculturated Apes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

INTRODUCTION

All animal species have unique characteristics and human beings are no excep-

tion. Perhaps most important, human beings have some unique cognitive skills.

The precise nature of these skills is unknown, but they must be such that they

enable a number of species-unique and easily observable behavioral practices,

including the following: (a) the creation and use of conventional symbols, includ-

ing linguistic symbols and their derivatives, such as written language and mathe-

matical symbols and notations; (b) the creation and use of complex tools and other

instrumental technologies; and (c) the creation of and participation in complex

social organizations and institutions. It is difficult to imagine a more fundamental

anthropological question than that of where these complex and species-unique

behavioral practices, and the cognitive skills that underlie them, came from.Recent research on human evolution has provided some important facts that

may help us to answer this most basic of questions. First, human beings shared acommon ancestor with their nearest primate relatives, Pan troglodytes and Pan

paniscus, a mere 6 million years ago—about the same time there existed a com-mon ancestor for horses and zebras, lions and tigers, and rats and mice (King &Wilson 1975). Second, for almost two thirds of this 6 million years, the humanlineage consisted of one or more species of Australopithicine, which in mostrecent research are characterized as ape-like in both brain size and behavior(Klein 1989). Third, although controversial, many anthropologists now believethat in the 2 million years of the existence of genus Homo, it has only been duringthe past several hundred thousand years, with the rise of something like modernhumans, that the unique aspects of human cognition have come into full bloom(Stringer & McKie 1996). What these new facts and interpretations establish isthe rapidity with which the species-unique aspects of human cognition must havearisen: within the past 6 million years for certain, within the past 2 million years inall likelihood, and within the past half million years according to some respect-able theories. The main point is that under none of these scenarios—especially thelast—has there been sufficient time for a large number of major cognitive adapta-tions (contra most of so-called evolutionary psychology) (Tooby & Cosmides1989, Pinker 1997). If we are searching for the origins of uniquely human cogni-tion, therefore, our search must be for some small difference that made a big dif-ference—some adaptation, or small set of adaptations, that changed the process ofprimate cognitive evolution in fundamental ways.

In my view there is only one candidate for this small difference that made a big

difference and that is human culture. Other primates and mammals are certainly

social, and some may even have social organizations for which it is useful to

apply the term culture (McGrew 1998). But human social organization is some-

thing else again, and this organization was, in my view, an integral part of the

process by which human cognition came to have many of its most distinctive

characteristics. That is, although the cognition of many mammalian and primate

species is influenced in important ways by their social environments, human cog-

nition, at least in its species-unique aspects, is actually socially constituted. In this

510 TOMASELLO

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

paper, I attempt to explicate this proposition more fully and to explore some of its

most important anthropological implications by systematically comparing the

social learning, social cognition, and cultural organization of human beings and

their nearest primate relatives.

PRIMATE AND HUMAN COGNITION

Human cognition is a species, in the literal meaning of the word, of primate cogni-tion. Tomasello & Call (1997) reviewed all of the most important studies of pri-mate cognition over the past century and established that a vast array of cognitiveskills are common to all primates, including humans. Thus, in their cognition ofthe physical world, all primate species remember “what” is “where” in their localenvironments, take novel detours and shortcuts in navigating through space (cog-nitive mapping), predict where food will be located in the future based on anumber of current cues, follow the visible and invisible displacements of objects(Piaget’s object permanence), categorize objects on the basis of perceptual simi-larities, understand relational categories and perform mental rotations of objectsin space, match (and perhaps add) small numerosities of objects independent ofspatial cues, and use creative strategies and perhaps insight in problem solving(sometimes in tool use). The major conclusion is that all primates live in basicallythe same sensory-motor world of permanent objects—and categories and rela-tions of permanent objects—arrayed in a representational space, and they all havesome insightful problem-solving skills (in some cases involving the making andusing of tools) to affect that sensory-motor world.

There are also many similarities in the way all primate species understand their

social worlds. Thus, all primate species recognize individuals in their social

groups; form direct relationships with other individuals based on such things as

kinship, friendship, and dominance rank; understand the third-party social rela-

tionships that hold among other members of their groups—again based on such

things as kinship, friendship, and dominance rank; predict the behavior of indi-

viduals based on a variety of social and physical cues, and sometimes novel

insights; use many types of social and communicative strategies to solve social

problems and so to out-compete groupmates for valued resources; cooperate with

conspecifics in problem-solving tasks and in forming social coalitions and alli-

ances; and engage in various forms of social learning, in which they learn valu-

able things from conspecifics. The major conclusion again is that all primates live

in basically the same type of social world, in which they individually recognize

conspecifics and appreciate both the vertical (dominance) and horizontal (affilia-

tive) relationships that hold among group members. They also have the ability to

predict the behavior of conspecifics in many situations based on a variety of cues

and insights, and in some cases to affect the behavior of groupmates via various

social and communicative strategies.So what makes human cognition different? What enables human beings to cre-

ate and use language and other symbols, to create and maintain complex instru-

mental technologies, and to create and maintain complex social organizations and

HUMAN ADAPTATION FOR CULTURE 511

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

institutions? The first and most obvious observation is that individual human

beings do not do any of these things. These are all collective cognitive products in

which human beings have in some way pooled their cognitive resources. If we

imagine the forbidden experiment in which a human child grows up on a desert

island, miraculously supplied with nutritional and emotional sustenance but in the

total absence of contact with other human beings, this child would not invent a

language or a complex technology or a complex social institution. Even if there

were a group of such abandoned children, it is unlikely that together, in their own

lifetime, they could invent anything resembling the range of material and sym-

bolic artifacts that characterizes even the least artifactually complex human soci-

ety. The reason that no single child or group of children could on their own in their

own lifetimes create any version of a modern human culture and its material and

symbolic artifacts is that human cultures are historical products built up over

many generations. Indeed, the most distinctive characteristic of human cultural

evolution as a process is the way that modifications to an artifact or a social prac-

tice made by one individual or group of individuals often spread within the group,

and then stay in place until some future individual or individuals make further

modifications—and these then stay in place until still further modifications are

made (Tomasello et al 1993a, Boesch & Tomasello 1998). This process of cumu-

lative cultural evolution works because of a kind of “ratchet effect”: Individual

and group inventions are mastered relatively faithfully by conspecifics, including

youngsters, which enables them to remain in their new and improved form within

the group until something better comes along.The major part of the ratchet in the cumulative cultural evolution of human

societies takes place during childhood. That is, each new generation of children

develops in the “ontogenetic niche” characteristic of its culture (including in

some cases explicitly pedagogical niches), mastering the artifacts and social prac-

tices that exist at that time. It is only because human children are so good at social

learning (and in some cases adults are so good at teaching) that an artifact or

social practice may conserve its form over many generations of stasis, until even-

tually a modification that group members find worthwhile is made and the cycle

starts anew. For this process to work, therefore, human beings not only need to be

inventive, they need to be good at preserving those inventions by imitatively

learning, and sometimes explicitly teaching, the inventions of others. This

process is more complex than it might seem at first glance, however. Imitative

learning does not just mean mimicking the surface structure of a poorly under-

stood behavior, the way a parrot mimics human speech, with no understanding of

its communicative significance, it also means reproducing an instrumental act

understood intentionally, that is reproducing not just the behavioral means but

also the intentional end for which the behavioral means was formulated. This

requires some specially adapted skills of social cognition.The main point is that unlike the young of any other primate species, human

children grow up in the midst of the accumulated wisdom of their social group, as

embodied in its material artifacts, symbolic artifacts, and conventional social

practices, and children are specifically adapted to appropriate this wisdom as

512 TOMASELLO

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

embodied in these forms. Although we have yet to explore the details, these facts

provide a sufficient explanation for the existence of many of the most distinctive

cognitive products that human beings produce. But there is more. As a result of

participating in social and communicative interactions with other persons under-

stood intentionally, human children also come to cognitively represent the world

in some uniquely powerful ways. The most important of these involves the use of

linguistic symbols that are both intersubjectively understood and perspectival, in

the sense that that a single item or situation may be construed linguistically in

many different ways. For example, a single event may be seen as X sold Y to Z or

Z bought Y from X; or a single object may be an apple, a fruit, or some food. Inter-

subjective and perspectival cognitive representations are unique in the animal

kingdom, and they enable human beings to deal with their worlds in some

uniquely flexible and powerful ways.To appreciate fully (a) the social cognitive skills necessary for children to par-

ticipate fully in their cultures and (b) the transforming effect of cultural participa-

tion on individual cognition, we must follow out key aspects of the human

ontogenetic scenario—and then compare it with the basic scenario of nonhuman

primate cognitive ontogeny.

THE ONTOGENY OF HUMAN CULTURAL LEARNING

Human children grow up in the ontogenetic niche of their culture, which, in asense, exists before they are born. But children also need to have some social cog-nitive skills if they are to exploit the preexisting cultural resources in a species-typical manner. These skills cannot be simply presupposed, as is often the case incultural psychology (e.g. Rogoff 1990, Shweder 1990). This point is most clearlydemonstrated by the unfortunate case of children with autism, the vast majority ofwhom lack the social cognitive skills necessary to participate fully in or to appro-priate the artifacts and social practices characteristic of those around them (Hob-son 1993, Baron-Cohen 1993). For typically developing children, the ontogeny ofthese social cognitive skills begins at the end of the first year of life.

Joint Attention

Six-month-old infants interact dyadically with objects, grasping and manipulat-ing them, and they interact dyadically with other people, expressing emotionsback-and-forth in a turn-taking sequence. But at approximately 9–12 months ofage, infants begin to engage in interactions that are triadic in the sense that theyinvolve the referential triangle of child, adult, and some outside entity to whichthey share attention. Thus, infants at this age begin to flexibly and reliably lookwhere adults are looking (gaze following), use adults as social reference points(social referencing), and act on objects in the way adults are acting on them (imi-tative learning)—in short, to “tune in” to the attention and behavior of adultstoward outside entities. At this same age, infants also begin to use communicativegestures to direct adult attention and behavior to outside entities in which they areinterested—in short, to get the adult to “tune in” to them. In many cases, several of

HUMAN ADAPTATION FOR CULTURE 513

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

these behaviors come together as the infant interacts with an adult in a relatively

extended bout of joint engagement with an object (Bakeman & Adamson 1984).

Most often the term joint attention has been used to characterize this whole com-

plex of triadic social skills and interactions (Moore & Dunham 1995), and it rep-

resents something of a revolution in the way infants relate to their worlds.Infants begin to engage in joint attentional interactions when they begin to

understand other persons as intentional agents (Tomasello 1995). Intentionalagents are animate beings with the power to control their spontaneous behavior,but they are more than that. Intentional agents also have goals and make activechoices among behavioral means for attaining those goals. It is important to notethat intentional agents also make active choices about what they pay attention toin pursuing those goals (for the argument that attention is intentional perception,see Gibson & Rader 1979). All of the specific joint attentional behaviors in whichinfants follow, direct, or share adult attention and behavior are not separate activi-ties or cognitive domains, they are simply different behavioral manifestations ofthis same underlying understanding of other persons as intentional agents. Strongsupport for this view comes from a recent study by Carpenter et al (1998b), whofollowed a group of infants longitudinally from 9 to 15 months of age and foundthat for any individual child these skills emerged together as a group, with somepredictable orderings among individual skills.

Imitative Learning

The social-cognitive revolution at 1 year of age, sets the stage for infants’ secondyear of life, in which they begin to imitatively learn the use of all kinds of tools,artifacts, and symbols. For example, in a study by Meltzoff (1988), 14-month-oldchildren observed an adult bend at the waist and touch its head to a panel, thusturning on a light. They followed suit. Infants engaged in this somewhat unusualand awkward behavior, even though it would have been easier and more naturalfor them simply to push the panel with their hand. One interpretation of thisbehavior is that infants understood that (a) the adult had the goal of illuminating

the light and then chose one means for doing so, from among other possible

means, and (b) if they had the same goal, they could choose the same means. Cul-

tural learning of this type thus relies fundamentally on infants’ tendency to iden-

tify with adults, and on their ability to distinguish in the actions of others the

underlying goal and the different means that might be used to achieve it. This

interpretation is supported by the more recent finding of Meltzoff (1995) that 18-

month-old children also imitatively learn actions that adults intend to perform,

even if they are unsuccessful in doing so. Similarly, Carpenter et al (1998a) found

that 16-month-old infants will imitatively learn from a complex behavioral

sequence only those behaviors that appear intentional, ignoring those that appear

accidental. Young children do not just mimic the limb movements of other per-

sons, they attempt to reproduce other persons’ intended actions in the world.Although it is not obvious at first glance, something like this same imitative

learning process must happen if children are to learn the symbolic conventions of

their native language. Although it is often assumed that young children acquire

514 TOMASELLO

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

language as adults stop what they are doing, hold up objects, and name these

objects for them, this is empirically not the case. Linguistics lessons such as these

are (a) characteristic of only some parents in some cultures and (b) characteristic

of no parent in no culture for words other than concrete nouns and some actions,

i.e. no one names for children acts of “giving” or prepositional relationships such

as “on” or “for.” In general, for the vast majority of words in their language, chil-

dren must find a way to learn in the ongoing flow of social interaction, sometimes

from speech not even addressed to them (Brown 1999). In some recent experi-

ments, something of this process has been captured, as children learned words in

situations in which the adult was not specifically intending that they learn a word,

the referent was not perceptually available when the word was said, and there

were multiple potential referents in the situation that the child had to choose

among based on various kinds of adult social-pragmatic cues.

1. In the context of a finding game, an adult announced her intentions to “findthe toma” and then searched in a row of buckets all containing novel objects.Sometimes she found it in the first bucket searched, smiling and handing thechild the object. Sometimes, however, she had to search longer, rejecting un-wanted objects by scowling at them and replacing them in their buckets untilshe found the one she wanted (again indicated by a smile and the termina-tion of search). Children learned the new word for the object the adult in-tended to find regardless of whether or how many objects were rejectedduring the search process (Tomasello & Barton 1994).

2. Also in the context of a finding game, an adult had the child find four differ-ent objects in four different hiding places, one of which was a distinctive toybarn. Once the child had learned which objects went in which places, theadult announced her intention to “find the gazzer.” She then went to the toybarn, but it turned out to be “locked.” She thus frowned at the barn and thenproceeded to another hiding place, saying “Let’s see what else we can find”(taking out an object with a smile). Later, children demonstrated that theyhad learned “gazzer” for the object they knew the experimenter wanted inthe barn—even though they never saw the object after they heard the newword, and even though the adult had frowned at the barn and smiled at a dis-tractor object (Akhtar & Tomasello 1996, Tomasello et al 1996).

3. An adult announced her intention to “dax Mickey Mouse” and then pro-ceeded to perform one action accidentally and another intentionally (orsometimes in reverse order). Children learned the word for the intentionalnot the accidental action regardless of which came first in the sequence(Tomasello & Barton 1994).

Tomasello et al (1993a) called this kind of imitative learning cultural learning

because the child is not just learning things from other persons, she is also learn-

ing things through them—in the sense that she must know something of the

adult’s perspective on a situation to learn the active use of this same intentional

act. The adult in the above scenarios is not just moving and picking up objects ran-

domly, she is searching for an object and the child must know this in order to

HUMAN ADAPTATION FOR CULTURE 515

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

make enough sense of her behavior to connect the new word to the adult’s

intended referent. The main theoretical point is that an organism can engage in

cultural learning of this type only when it understands others as intentional

agents, like the self, who have a perspective on the world that can be followed

into, directed, and shared. Indeed, a strong argument can be made that children

can only understand a symbolic convention in the first place if they understand

their communicative partner as an intentional agent with whom one may share

attention—since a linguistic symbol is nothing other than a marker for an inter-

subjectively shared understanding of a situation (Tomasello 1995, 1998, 1999).

As a point of comparison, children with autism do not understand other persons as

intentional agents, or they do so to only an imperfect degree, and so they do

poorly at imitative learning of intentional actions in general (Smith & Bryson

1994): Only half of them ever learn any language at all, and those who do learn

some language do poorly in word-learning situations such as those just described

(Baron-Cohen et al 1997). As we see below, nonhuman primates are not very

human-like in these kinds of social-cognitive and cultural learning skills either.

Linguistic Symbols and Cognitive Representation

One of the most interesting things about the process of language acquisition isthat the adults from whom the child is learning went through the same process ear-lier in their lives, and across generations the symbolic artifacts that comprise Eng-lish or Turkish, or whatever language, accumulate modifications as new linguisticforms are created by grammaticization and other processes of language change(e.g. Traugott & Heine 1991). Thus, today’s child is learning the whole histori-cally derived conglomeration. Consequently, when the child learns the conven-tional use of these well-traveled symbols, what she is learning is the ways herforbears in the culture found useful for manipulating the attention of others in thepast. And because the peoples of a culture, as they move through historical time,evolve many and varied purposes for manipulating the attention of one another(and because they need to do this in many different types of discourse situations),today’s child is faced with a whole panoply of different linguistic symbols andconstructions that embody many different attentional construals of any givensituation. As a sampling, languages embody attentional construals based on thefollowing (for more specifices, see Langacker 1987): generality-specificity(”thing,” “furniture,” “chair,” “desk chair”); perspective (”chase-flee,” “buy-sell,” “come-go,” “borrow-lend”); function (”father,” “lawyer,” “man,” “Ameri-can”) (”coast,” “shore,” “beach”). There are many more specific perspectives thatarise in grammatical combinations of various sorts: “She smashed the vase” ver-sus “The vase was smashed.” It is at about 18 months of age that children firstbegin to predicate multiple things about objects to which they and the adult arejointly attending, for example by saying that this ball is either “wet” or “big” or“mine”—all about one and the same object (Reed 1995; Tomasello 1988, 1995).

Consequently, as a young child internalizes a linguistic symbol, as she learns

the human perspective embodied in that symbol, she cognitively represents not

just perceptual or motoric aspects of a situation but also one way, among other

516 TOMASELLO

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

ways of which she is also aware, that the current situation may be attentionally

construed by “us,” the users of the symbol. The way that human beings use lin-

guistic symbols thus creates a clear break with straightforward perceptual or

sensory-motor cognitive representations. It is true that a prelinguistic child, or a

nonhuman primate, may construe situations in more than one way: One time a

conspecific is a friend and the next time an enemy; one time a tree is for climbing

to avoid predators and the next time it is for making a nest in. In these different

interactions with the same entity, the individual is deploying its attention differ-

entially, depending on its goal at that moment. But shifting attention sequentially

in this manner as a function of goal is not the same thing as knowing simultane-

ously a number of different ways in which something might be construed—in

effect, imagining at the same time a number of different possible goals and their

implications for attention. An individual language user looks at a tree and, before

drawing the attention of her interlocutor to that tree, must decide, based on her

assessment of the listener’s current knowledge and expectations, whether to say

“that tree over there,” “it,” “the oak,” “that hundred-year-old oak,” “the tree,”

“The bagswing tree,” “that thing in the front yard,” “the ornament,” “the embar-

rassment,” or any of a number of other expressions. She must decide whether the

tree is in, is standing in, is growing in, was placed in, or is flourishing in the front

yard. And these decisions are not made on the basis of the speaker’s direct goal

with respect to the object or activity involved, but rather they are made on the

basis of her goal with respect to the listener’s interest and attention to that object

or activity. This means that the speaker knows that the listener shares with her

these same choices for construal—again, all available simultaneously. Indeed,

the fact is that while she is speaking, the speaker is constantly monitoring the lis-

tener’s attentional status (and vice versa), which means that both participants in a

conversation are always aware that there are at least their two actual perspectives

on a situation, as well as many more that are symbolizable in currently unused

symbols and constructions.The point is not just that linguistic symbols provide handy tags for human con-

cepts, or even that they influence or determine the shape of those concepts, though

they do both of these things. The point is that the intersubjectivity of human lin-

guistic symbols—and their perspectival nature as one offshoot of this intersubjec-

tivity—means that linguistic symbols do not represent the world directly, in the

manner of perceptual or sensory-motor representations, but rather they are used

by people to induce others to construe certain perceptual/conceptual situations—

to attend to them—in one way rather than in another. This breaks symbols away

from the sensory-motor world of objects in space and puts them instead in the

realm of the human ability to view the world in whatever way is convenient for the

communicative purpose at hand. The most important point in the current context

is that as children participate in these communicative exchanges, they internalize,

in something like the way Vygotsky (1978) envisioned the process, the perspec-

tives of other persons. The internalization process is not something mystical or

unanalyzable, as it is sometimes characterized, but rather it is just the normal

process of imitative learning when linguistic (and perhaps other communicative)

HUMAN ADAPTATION FOR CULTURE 517

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

conventions are involved. That is, in imitatively learning a linguistic symbol from

an adult, the child comprehends that by using a particular symbol she intends for

another to pay attention to some specific aspect of their shared experience. When

the child attempts to appropriate the use of this communicative convention for her

own use, she must reverse roles: If others wish her to focus on this same aspect of

reality, they must use that same symbol toward her. This learning process is

indeed what creates the communicative convention in the first place, in the sense

that it initiates the child into the convention. Because linguistic symbols are per-

spectival, i.e. used to focus the attention of others on specific aspects of situations

as opposed to other aspects, if the child is to use the symbol in its conventionally

appropriate manner she must understand something of the adult’s perspective. It

is in this sense and only in this sense that internalization involves a special form of

social learning—cultural learning—in which the child internalizes the perspec-

tive of another person.Some of the effects of operating with symbols of this type are obvious, in terms

of flexibility and relative freedom from perception. But some are more far reach-

ing and quite unexpected, in the sense that they give children truly new ways of

conceptualizing things, such as treating objects as actions (“he porched the news-

paper”), actions as objects (“skiing is fun”), and all kinds of metaphorical constru-

als of things (“love is a journey”) (Lakoff & Johnson 1980, Lakoff 1987, Johnson

1987). These new ways of conceptualizing and thinking result from the accumu-

lated effects of engaging in linguistic communication with other persons for some

years during early cognitive development. More extended bouts of discourse

interaction with other persons also create opportunities for explicitly exploring

and comparing differering verbally expressed perspectives on situations. Perhaps

of special importance are discourse interactions in which the communicative part-

ner provides a verbally expressed perspective on the child’s previous verbally

expressed perspective, since in this case the internalization of the other’s perspec-

tive helps to create children’s ability to self-regulate, self-monitor, and reflect on

their own cognition (Vygotsky 1978).

NONHUMAN PRIMATE CULTURE

McGrew (1998) claims that nonhuman primates engage in social activities that

are best characterized as cultural in that they share all the essential features of

human culture. I agree with this (Tomasello, 1990, 1994, 1996a). Nevertheless, at

the same time I insist that human culture has, in addition, some unique character-

istics (as may the cultures of other primate species). The most important of these,

at the macro-level, is the fact that many human cultural traditions and artifacts

accumulate modifications over time, whereas this does not seem to be the case for

nonhuman primate cultural traditions (Tomasello et al 1993a, Tomasello & Call

1997). The reason for this difference resides in the micro-level processes by

which individuals learn things from and through one another, and as just elabo-

rated, it may be the case that in their early ontogenies, human beings do this in

some species-unique ways. To see whether these skills are indeed unique to

518 TOMASELLO

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

humans, we should look briefly at nonhuman primate culture, nonhuman primate

social learning, and nonhuman primate social cognition.

Japanese Macaque Potato Washing

The most often-cited case of nonhuman primate culture is the case of Japanese

macaque potato washing (Kawamura 1959, Kawai 1965). In 1953, an 18-month-

old female named Imo was observed to take pieces of sweet potato, given to her

and the rest of the troop by researchers, and to wash the sand off of them in some

nearby water (at first a stream and then the ocean). About 3 months after she

began to wash her potatoes, the practice was observed in Imo’s mother and two of

her playmates (and then their mothers). During the next 2 years, seven other

youngsters also began to wash potatoes, and within 3 years of Imo’s first potato

washing approximately 40% of the troop was doing the same. It was thought sig-

nificant that it was Imo’s close associates who learned the behavior first, and their

associates directly after, in that it suggested that the means of propagation of this

behavior was some form of imitation, in which one individual actually copied the

behavior of another.The interpretation of these observations in terms of culture and imitation has

two main problems, however. The first is that potato washing is much lessunusual a behavior for monkeys than was originally thought. Brushing sand offfood turns out to be something that many monkeys do naturally, and indeed thishad been observed in the Koshima monkeys prior to the emergence of washing. Itis thus not surprising that potato washing was also observed in four other troops ofhuman-provisioned Japanese macaques soon after the Koshima observations(Kawai 1965), which implies that at least four individuals learned on their own.Also, in captivity, individuals of other monkey species learn quite rapidly on theirown to wash their food when provided with sandy fruits and bowls of water(Visalberghi & Fragaszy 1990). The second problem has to do with the pattern ofthe spread of potato washing behavior within the group. The spread of the behav-ior was relatively slow, with an average time of over 2 years for acquisition by allthe members of the group who learned it (Galef 1992). Moreover, the rate ofspread did not increase as the number of users increased. If the mechanism oftransmission was imitation, an increase in the rate of propagation would beexpected as more demonstrators became available for observation over time. Incontrast, if processes of individual learning were at work, a slower and steadierrate of spread would be expected—which was in fact observed. The fact thatImo’s friends and relatives were first to learn the behavior may be due to the factthat friends and relative stay close to one another, and thus Imo’s friends verylikely went near the water more often during feeding than other group members,increasing their chances for individual discovery.

Chimpanzee Tool Use

Perhaps the best examples to examine in the current context are humans’ closest

primate relatives, the chimpanzees—especially with regard to tool use and ges-

HUMAN ADAPTATION FOR CULTURE 519

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

tural communication, the two behaviors for which cultural transmission has been

most often claimed. First, there are a number of population-specific tool use tradi-

tions that have been documented for different chimpanzee communities, for

example termite-fishing, ant-fishing, ant-dipping, nut-cracking, and leaf-

sponging (for a review, see McGrew 1992). Sometimes the “same” tradition even

shows variability between groups. For instance, members of the Kasakela com-

munity at Gombe (as well as some other groups elsewhere) fish for termites by

probing termite mounds with small, thin sticks, whereas in other parts of Africa

there are chimpanzees who perforate termite mounds with large sticks and

attempt to scoop up the insects by the handful. One possible explanation is that the

chimpanzees in the western parts of Africa are able to destroy termite mounds

with large sticks because the mounds are soft from much rain, whereas in the east

they cannot use this strategy because the mounds are too hard. Each individual

thus reinvents the wheel for itself, with population differences due to the different

local ecologies of the different groups—so-called environmental shaping.Although environmental shaping is likely a part of the explanations for group

differences of behavior for all species, experimental studies have demonstratedthat more than this is going on in chimpanzee culture (see also Boesch et al 1994).Tomasello (1996a) reviewed all the experimental evidence on chimpanzee imita-tive learning of tool use (a total of five studies) and concluded that chimpanzeesare good at learning about the dynamic affordances of objects they see beingmanipulated by others, but they are not skillful at learning from others a newbehavioral strategy per se. For example, if a mother rolls over a log and eats theinsects underneath, her child will likely follow suit. This is simply because thechild learned from the mother’s act that there are insects under the log—a fact shedid not know and very likely would not have discovered on her own. But she didnot learn how to roll over a log or to eat insects; these are things she already knewhow to do or could learn how to do on her own. (Thus, the youngster would havelearned the same thing if the wind, rather than its mother, had caused the log toroll over and expose the ants.) This is what has been called emulation learningbecause it is learning that focuses on the environmental events involved—theresults or changes of state in the environment that the other produced—rather thanon the actions that produced those results (Tomasello 1990, 1996a).

Emulation learning is a very intelligent and creative learning process that, in

some circumstances, is a more adaptive strategy than imitative learning. For

example, Nagell et al (1993) presented chimpanzees and 2-year-old human chil-

dren with a rake-like tool and an out-of-reach object. The tool was such that it

could be used in either of two ways, leading to the same end-result of obtaining

the object. For each species, one group of subjects observed a demonstrator

employ one method of tool use (less efficient) and another group of subjects

observed another method of tool use (more efficient). The result was that whereas

human children in general copied the method of the demonstrator in each of the

two observation conditions (imitative learning), chimpanzees used the same

method or methods to obtain the object no matter which demonstration they

observed (emulation learning). The interesting point is that many children insisted

520 TOMASELLO

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

on this reproduction of adult behavior even in the case of the less-efficient

method, leading to less successful performance than the chimpanzees in this con-

dition. Imitative learning is not a “higher” or “more intelligent” learning strategy

than emulation learning; it is simply a more social strategy—which, in some cir-

cumstances and for some behaviors, has some advantages. This emulation learn-

ing explanation also applies to other studies of chimpanzee social learning of tool

use activities, such as those by Whiten et al (1996) and Russon & Galdikas

(1993).Chimpanzees are intelligent and creative in using tools and understanding

changes in the environment brought about by the tool use of others, but they do

not seem to understand the instrumental behavior of conspecifics the same way as

do humans. For humans, the goal or intention of the demonstrator is a central part

of what they perceive, and indeed the goal is understood as something separate

from the various behavioral means that may be used to accomplish the goal. An

observer’s ability to separate goal and means serves to highlight for herself the

demonstrator’s method or strategy of tool use as an independent entity—the

behavior she is using in an attempt to accomplish the goal, given the possibility of

other means of accomplishing it. In the absence of this ability to understand goal

and behavioral means as separable in the actions of others, chimpanzee observers

focus on the changes of state (including changes of spatial position) of the objects

involved during the demonstration, with the motions of the demonstrator being, in

effect, just other motions. The intentional states of the demonstrator, and thus her

behavioral methods as distinct behavioral entities, are simply not a part of their

experience.

Chimpanzee Gestural Communication

The other well-known case is the gestural communication of chimpanzees, for

which there are also some population-specific behaviors (Goodall 1986,

Tomasello 1990). There is one reasonably well-documented example from the

wild. Nishida (1980) reported “leaf-clipping” in the Mahale K group of chimpan-

zees—thought to be unique to that group but later observed by Sugiyama (1981)

in another group across the continent. The reporting of data for individuals in

these studies showed that there were marked individual differences within the

groups in how (toward what end) the signal was used, for example for sexual

solicitation, aggression toward groupmates, or aggression toward humans. One

hypothesis is that after one individual used leaf-clipping to make noise (the tear-

ing of the rigid dead leaves makes a very loud noise), others learned via emulation

to make the same noise (i.e. they learned the affordances of the leaf). This had dif-

ferent attention-getting effects on conspecifics in the different groups, however,

and these were then learned as contingencies. The fact that leaf-clipping has been

observed in more than one group, who have had no opportunity to observe one

another, raises the possibility of some such process.This possibility is also supported by studies with captive chimpanzees. In

ongoing studies of the gestural signaling of a captive colony of chimpanzees,

Tomasello and colleagues have asked whether youngsters acquire their gestural

HUMAN ADAPTATION FOR CULTURE 521

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

signals by imitative learning or by a process of ontogenetic ritualization

(Tomasello et al 1985, 1989, 1994, 1997). In ontogenetic ritualization, a commu-

nicatory signal is created by two organisms shaping each other’s behavior in

repeated instances of a social interaction. For example, an infant may initiate

nursing by going directly for the mother’s nipple, perhaps grabbing and moving

her arm in the process. In some future encounter the mother might anticipate the

infant’s impending behavioral efforts at the first touch of her arm, and so become

receptive at that point, leading the infant on some future occasion still to abbrevi-

ate its behavior to a touch on the arm while waiting for a response (“arm-touch” as

a so-called intention movement). Note that there is no hint here that one individ-

ual is seeking to reproduce the behavior of another; there is only reciprocal social

interaction over repeated encounters that results eventually in a communicative

signal. This is presumably the way that most human infants learn the “arms-over-

head” gesture to request that adults pick them up (Lock 1978).All of the available evidence suggests that ontogenetic ritualization, not imita-

tive learning, is responsible for chimpanzees’ acquisition of communicative ges-

tures. First, there are a number of idiosyncratic signals that are used by only one

individual (see also Goodall 1986). These signals could not have been learned by

imitative processes and so must have been individually invented and ritualized.

Second, longitudinal analyses have revealed quite clearly, by both qualitative and

quantitative comparisons, that there is much individuality in the use of gestures,

with much individual variability both within and across generations, which sug-

gesting something other than imitative learning. It is also important that the ges-

tures that are shared by many youngsters are gestures that are also used frequently

by captive youngsters raised in peer groups with no opportunity to observe older

conspecifics. Finally, in an experimental study, Tomasello et al (1997) removed

an individual from the group and taught her two different arbitrary signals by

means of which she obtained desired food from a human. When she was then

returned to the group and used these same gestures to obtain food from a human in

full view of other group members, there was not one instance of another individ-

ual reproducing either of the new gestures.The clear conclusion is that chimpanzee youngsters acquire the majority, if not

the totality, of their gestures by individually ritualizing them with one another.

The explanation for this learning process is analogous to the explanation for emu-

lation learning in the case of tool use. Like emulation learning, ontogenetic ritu-

alization does not require individuals to understand the behavior of others as

separable into means and goals in the same way as does imitative learning. Imita-

tively learning an arm-touch as a solicitation for nursing would require that an

infant observe another infant using an arm-touch and know what goal it was pur-

suing (viz. nursing), so that when it had the same goal it could use the same behav-

ioral means (viz. arm-touch). Ritualizing an arm-touch, on the other hand, only

requires the infant to anticipate the future behavior of a conspecific in a context in

which it (the infant) already has the goal of nursing. Ontogenetic ritualization is

thus, like emulation learning, an intelligent and creative social learning process

that is important in all social species, including humans. But it is not a learning

522 TOMASELLO

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

process by means of which individuals attempt to reproduce the behavioral strate-gies of others.

Nonhuman Primate Social Learning and Social Cognition

Chimpanzee tool use and gestural communication thus provide us with two verydifferent sources of evidence about nonhuman primate social learning. In the caseof tool use, it is likely that chimpanzees acquire the tool use skills they areexposed to by a process of emulation learning. In the case of gestural signals, it islikely that they acquire their communicative gestures through a process of onto-genetic ritualization. Both emulation learning and ontogenetic ritualizationrequire skills of cognition and social learning, each in its own way, but neitherrequires skills of imitative learning, in which the learner comprehends both thedemonstrator’s goal and the strategy she is using to pursue that goal—and then insome way aligns this goal and strategy with her own. Indeed, emulation learningand ontogenetic ritualization are precisely the kinds of social learning one wouldexpect of organisms that are intelligent and quick to learn, but that do not under-stand others as intentional agents with whom they can align themselves.

The other main process involved in cultural transmission as traditionallydefined is teaching. Whereas social learning comes from the “bottom up,” asignorant or unskilled individuals seek to become more knowledgeable or skilled,teaching comes from the “top down,” as knowledgeable or skilled individualsseek to impart knowledge or skills to others. The problem in this case is that thereare few systematic studies of teaching in nonhuman primates. The most thoroughstudy is that of Boesch (1991), in which chimpanzee mothers and infants wereobserved in the context of tool use (nut cracking). Boesch discovered that amother does a number of things that serve to facilitate the infant’s activities withthe tool and nuts, such as leaving the tools idle while she goes to gather more nuts(which she would not do if another adult were present). But the interpretation ofthe mother’s intention in such cases is far from straightforward. Moreover, in thecategory of “active instruction,” in which the mother appears to be activelyattempting to instruct her child, Boesch observed only two possible instances(over many years of observation). These two instances are also difficult to inter-pret, in the sense that the mother may or may not have had the goal of helping theyoungster learn to use the tool. On the other hand, although there is much variabil-ity across different human societies, adult humans in all cultures actively instructtheir young on a regular basis in one way or another (Kruger & Tomasello 1996).Along with imitative learning, the process of active instruction is likely crucial tothe uniquely human pattern of cumulative cultural evolution as well.

It should be acknowledged, of course, that this way of viewing nonhuman pri-

mate culture and social learning is not the only way of viewing them, and indeed

researchers such as McGrew (1998) and Boesch (1999) would likely disagree

with many of the current conclusions. Indeed, there are also a number of widely

publicized studies purportedly demonstrating that nonhuman primates have theo-

ries of mind, deceive their conspecifics, and engage in many other kinds of “mind-

reading” activities (e.g. Byrne 1995). If these studies represented the true picture

HUMAN ADAPTATION FOR CULTURE 523

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

of nonhuman primate social cognition, it would be a complete mystery why they

did not engage in more powerful forms of cultural learning. However, there are in

actuality surprisingly few of these studies, and as in all scientific paradigms, a few

studies is not enough to settle opinion. My own interpretation of these studies is

that they represent cases in which nonhuman primates have acquired clever

strategies to manipulate the behavior, not the mental states, of others (for detailed

analyses of all the relevant studies, see Tomasello 1996b; Tomasello & Call

1994, 1997). If they are not dealing with mental or intentional states, but only

behavior, nonhuman primates will not engage in the kinds of cultural learning

that lead to the ratchet effect—and that thus create human-like cultural niches full

of symbolic artifacts and instructional formats for the cognitive ontogeny of their

offspring.With regard to nonhuman primate culture in particular, Boyd & Richerson

(1996), in a paper entitled “Why culture is common, but cultural evolution is

rare,” hypothesize that humans and other animals both engage in the same kinds

of social and imitative learning but with a quantitative difference. Thus, chimpan-

zees may have some imitative learning abilities, but they may display them less

consistently or in a narrower range of contexts than do humans—or it may even be

that only some individuals have these skills. These authors go on to make a quan-

titatively based evolutionary argument that this rarity itself can make cultural

evolution of the cumulative type impossible. So it might be that a quantitative dif-

ference in social learning leads to a qualitative difference in the nature of cultural

traditions, and in particular how they change and evolve over time. Although

there are currently no easy ways to quantify the frequency of imitative learning in

the societies of different species, this hypothesis is intriguing because it posits the

existence of the kind of variation that might have characterized the human-

chimpanzee common ancestor and with which subsequent evolutionary processes

might have worked.

Enculturated Apes

It may be objected that there are a number of convincing observations of chim-

panzee imitation in the literature, and indeed there are a few. It is interesting, how-

ever, that basically all the clear cases in the exhaustive review of Whiten & Ham

(1992) concern chimpanzees that have had extensive amounts of human contact.

In many cases, this has taken the form of intentional instruction involving human

encouragement of behavior and attention, and even direct reinforcement for imi-

tation for many months [e.g. 7 months of training in the case of Hayes & Hayes

(1952) and 4 months of training in the case of Whiten & Custance (1996)]. This

raises the possibility that imitative learning skills may be influenced, or even

enabled, by certain kinds of social interaction during early ontogeny.Confirmation for this point of view is provided in a study by Tomasello et al

(1993b). It compared the imitative learning abilities of mother-reared captive

chimpanzees, enculturated chimpanzees (raised like human children and exposed

to a language-like system of communication), and 2-year-old human children.

Each subject was shown 24 different and novel actions on objects, and each sub-

524 TOMASELLO

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

ject’s behavior on each trial was scored as to whether it successfully reproduced

(a) the end result of the demonstrated action and/or (b) the behavioral means used

by the demonstrator. The major result was that the mother-reared chimpanzees

reproduced both the end and the means of the novel actions (i.e. imitatively

learned them) hardly at all. In contrast, the enculturated chimpanzees and the

human children imitatively learned the novel actions much more frequently, and

they did not differ from one another in this learning. Interesting corroboration for

this latter finding is the fact that earlier in their ontogeny, these same enculturated

chimpanzees seemed to learn many of their human-like symbols by means of imi-

tative learning (Savage-Rumbaugh 1990).For the issue of chimpanzee culture in the wild, these results raise an important

question. Which group of captive chimpanzees is more representative of chim-

panzees in their natural habitats: mother-reared or enculturated? Are enculturated

chimpanzees simply displaying more species-typical imitative learning skills

because their more enriched rearing conditions more closely resemble those of

wild chimpanzees than do the impoverished rearing conditions of other captive

chimpanzees? Or might it be the case that the human-like socialization process

experienced by enculturated chimpanzees differs significantly from the natural

state and, in effect, helps to create a set of species-atypical abilities more similar

to those of humans? There can be no definitive answer to these questions at this

time, but one possibility is that a human-like sociocultural environment is an

essential component in the development of human-like social-cognitive and imi-

tative learning skills, no matter the species. That is, this is true not only for chim-

panzees but also for human beings—a human child raised in an environment

lacking intentional interactions and other cultural characteristics in all likelihood

would also not develop human-like skills of imitative learning.The hypothesis is thus that the understanding of the intentions of others, neces-

sary for reproducing another’s behavioral strategies, develops in, and only in, the

context of certain kinds of intentional interactions with others (Tomasello et al

1993a). More specifically, to come to understand others in terms of their inten-

tions requires that the learner him- or herself be treated as an intentional agent in

which another organism encourages attention to and specific behaviors toward

some object of mutual interest—often reinforcing in some manner the learner’s

successful attempts in this direction (Call & Tomasello 1994, 1996). Such inter-

actions are not sufficient, of course, as many animals are subjected to all kinds of

human interaction, and even direct instruction, without developing human-like

skills of imitative learning (and the same is true of human children with autism).

The important point for current purposes is that in terms of these dimensions of

social interaction, captive chimpanzees raised by conspecifics are a better model

for wild chimpanzees than are chimpanzees raised in human-like cultural envi-

ronments—since wild chimpanzees receive little in the way of direct instruction

from conspecifics.A corollary hypothesis is thus that the learning skills that chimpanzees develop

in the wild in the absence of human interaction (i.e. skills involving individual

learning supplemented by emulation learning and ritualization) are sufficient to

HUMAN ADAPTATION FOR CULTURE 525

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

create and maintain their species-typical cultural activities, but they are not suf-

ficient to create and maintain human-like cultural activities displaying the

ratchet effect and cumulative cultural evolution. The fact that chimpanzees and

other great apes raised from an early age and for many years in human-like cul-

tural environments may develop some aspects of human social cognition and cul-

tural learning demonstrates the power of cultural processes in ontogeny in a

particularly dramatic way. The effect of cultural environments on nonhuman pri-

mate cognitive development is thus a question that deserves more empirical

investigation.

CONCLUSIONS

It is easy to observe a human behavior and posit a specific gene for that behavior,with no research into the genetics of the situation. Many scholars and popularistsof biological and social sciences make their living doing just that. But when wehave behaviors that are unique to a species, we have serious time constraints onsuch hypotheses, and so the positing of a large number of significant geneticevents becomes highly implausible. In the case of humans, the time frame for theemergence of their unique cognitive skills is almost certainly in the range of 2million to 0.3 million years ago—with my own theoretical bias being toward thesmaller of those figures. But the genetic event that happened at that time was notan everyday genetic event. It was not an everyday genetic event because it did notjust change one relatively isolated characteristic, it changed the nature of primatesocial cognition, which changed the social-cultural transmission process charac-teristic of primates, which led to a series of cascading sociological and psycho-logical events in historical time. The new form of social cognition that started theentire process involved understanding other persons as intentional agents like theself, and the new process of social-cultural transmission involved several formsof cultural learning, the first and most important being imitative learning. Thesenew forms of cultural learning created the possibility of a kind of ratchet effect inwhich human beings not only pooled their cognitive resources contemporane-ously, they also built on one another’s cognitive inventions over time. This newform of cultural evolution thus created artifacts and social practices with a his-tory, so that each new generation of children grew up in something like the accu-mulated wisdom of their entire social group, past and present.

And so, while not denying that a significant genetic event happened in human

cognitive evolution, probably fairly recently, I would deny that this event speci-

fied the detailed outcomes we see in adult humans today. In my view, that genetic

event merely opened the way for some new social and cultural processes that

then, with no further genetic events, created many, if not all, of the most interest-

ing and distinctive characteristics of human cognition. Perhaps of special impor-

tance was a new form of perspectively based cognitive representation that

emerged when children began to learn and use linguistic symbols, evolved over

historical time for inducing others to construe certain situations in certain ways,

which necessitated an internalization and representation of the different perspec-

526 TOMASELLO

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

tives of other persons. This story is not as simple as the “genes create all novel-ties” story, but it accords better with data from both anthropological anddevelopmental psychological investigations. Modern human adult cognition is aresult not just of processes of biological evolution, but also of other processes thathuman biological evolution made possible in both cultural historical time andindividual ontogenetic time.

Visit the Annual Reviews home page at www.AnnualReviews.org.

HUMAN ADAPTATION FOR CULTURE 527

LITERATURE CITED

Akhtar N, Tomasello M. 1996. Twenty-four

month old children learn words for absent

objects and actions. Br. J. Dev. Psychol.

14:79–93

Bakeman R, Adamson L. 1984. Coordinating

attention to people and objects in mother-

infant and peer-infant interactions. Child

Dev. 55:1278–89

Baron-Cohen S. 1993. From attention-goal

psychology to belief-desire psychology:

the development of a theory of mind and its

dysfunction. In Understanding Other

Minds: Perspectives from Autism, ed. S

Baron-Cohen, H Tager-Flusberg, DJ Co-

hen, pp. 59–82. New York: Oxford Univ.

Press

Baron-Cohen S, Baldwin D, Crowson M.

1997. Do children with autism use the

speaker’s direction of gaze to crack the

code of language? Child Dev. 68:48–57

Boesch C. 1991. Teaching among wild chim-

panzees. Anim. Behav. 41:530–32

Bosech C. 1999. The Chimpanzees of the Tai

Forest. Oxford, UK: Oxford Univ. Press.

In press

Boesch C, Marchesi P, Marchesi N, Fruth B,

Joulian F. 1994. Is nut cracking in wild

chimpanzees a cultural behavior? J. Hum.

Evol. 26:325–38

Boesch C, Tomasello M. 1998. Chimpanzee

and human culture. Contemp. Anthropol.

39:591–611

Bowerman M, Levinson S, eds. 1999. Lan-

guage Acquisition and Conceptual Devel-

opment. Cambridge, UK: Cambridge Univ.

Press

Boyd R, Richerson P. 1996. Why culture is

common but cultural evolution is rare.

Proc. Br. Acad. 88:77–93

Brown P. 1999. The conversational context for

language acquisition: a Tzeltal (Mayan)

case study. See Bowerman & Levinson

1999. In press

Byrne R. 1995. The thinking ape: Evolution-

ary Origins of Intelligence. Oxford, UK:

Oxford Univ. Press

Call J, Tomasello M. 1994. Production and

comprehension of referential pointing by

orangutans (Pongo pygmaeus). J. Comp.

Psychol. 108:307–17

Call J, Tomasello M. 1996. The role of humans

in the cognitive development of apes. In

Reaching into Thought: The Minds of the

Great Apes, ed. A Russon, pp. 371–403.

Cambridge, UK: Cambridge Univ. Press

Carpenter M, Akhtar N, Tomasello M. 1998a.

Sixteen-month-old infants differentially

imitate intentional and accidental actions.

Infant Behav. Dev. 21:315–30

Carpenter M, Nagell K, Tomasello M. 1998b.

Social cognition, joint attention, and com-

municative competence from 9 to 15

months of age. Monogr. Soc. Res. Child

Dev. 63

Galef B. 1992. The question of animal culture.

Hum. Nat. 3:157–78

Gibson E, Rader N. 1979. Attention: the per-

ceiver as performer. In Attention and Cog-

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

528 TOMASELLO

nitive Development, ed. G Hale, M Lewis,

pp. 6–36. New York: Plenum

Goodall J. 1986. The Chimpanzees of Gombe.

Patterns of behavior. Cambridge, MA:

Harvard Univ. Press

Hayes KJ, Hayes C. 1952. Imitation in a home-

raised chimpanzee. J. Comp. Psychol. 45:

450–59

Heyes C, Galef B, eds. 1996. Social Learning

in Animals: The Roots of Culture. New

York: Academic

Hobson P. 1993. Autism and the Development

of Mind. Hillsdale, NJ: Erlbaum

Johnson M. 1987. The Body in the Mind. Chi-

cago: Univ. Chicago Press

Kawai M. 1965. Newly-acquired pre-cultural

behavior of the natural troop of Japanese

monkeys on Koshima Islet. Primates 6:

1–30

Kawamura S. 1959. The process of subcultural

propagation among Japanese macaques.

Primates 2:43–60

King M, Wilson A. 1975. Evolution at two lev-

els in humans and chimpanzees. Science

188:107–16

Klein R. 1989. The Human Career: Human

Biological and Cultural Origins. Chicago:

Univ. Chicago Press

Kruger A, Tomasello M. 1996. Cultural learn-

ing and learning culture. In Handbook of

Education and Human Development: New

Models of Teaching, Learning, and School-

ing, ed. D Olson, pp. 169–87. London:

Blackwell

Lakoff G. 1987. Women, Fire, and Dangerous

Things: What Categories Reveal About the

Mind. Chicago: Univ. Chicago Press

Lakoff G, Johnson M. 1980. Metaphors We

Live By. Chicago: Univ. Chicago Press

Langacker R. 1987. Foundations of Cognitive

Grammar, Vol. 1. Stanford, CA: Stanford

Univ. Press

Locke A. 1978. The emergence of language. In

Action, Gesture, and Symbol: the Emer-

gence of Language, ed. A Locke, pp. 3–20.

New York: Academic

McGrew WC. 1992. Chimpanzee Material

Culture. Cambridge, UK: Cambridge

Univ. Press

McGrew WC. 1998. Culture in nonhuman pri-

mates? Annu. Rev. Anthropol. 27:301–28

Meltzoff A. 1988. Infant imitation after a one

week delay: long term memory for novel

acts and multiple stimuli. Dev. Psychol.

24:470–76

Meltzoff A. 1995. Understanding the inten-

tions of others: re-enactment of intended

acts by 18-month-old children. Dev. Psy-

chol. 31:838–50

Moore C, Dunham P. 1995. Joint Attention: Its

Origins and Role in Development. Hills-

dale, NJ: Erlbaum

Nagell K, Olguin K, Tomasello M. 1993. Proc-

esses of social learning in the tool use of

chimpanzees (Pan troglodytes) and human

children (Homo sapiens). J. Comp. Psy-

chol. 107:174–86

Nishida T. 1980. The leaf-clipping display: a

newly-discovered expressive gesture in

wild chimpanzees. J. Hum. Evol. 9:117–28

Pinker S. 1997. How the Mind Works. London:

Lane

Reed E. 1995. The ecological approach to lan-

guage development. Lang. Commun. 15:

1–29

Rogoff B. 1990. Apprenticeship in Thinking.

Oxford, UK: Oxford Univ. Press

Russon A, Galdikas B. 1993. Imitation in ex-

captive orangutans. J. Comp. Psychol. 107:

147–61

Savage-Rumbaugh S. 1990. Language as a

cause-effect communication system. Phi-

los. Psychol. 3:55–76

Shweder R. 1990. Cultural psychology—

What is it? In Cultural Psychology, ed. J

Stigler, R Shweder, G Herdt, pp. 1–46.

Cambridge, UK: Cambridge Univ. Press

Smith IM, Bryson SE. 1994. Imitation and ac-

tion in autism: a critical review. Psychol.

Bull. 116:259–73

Stringer C, McKie R. 1996. African Exodus:

The Origins of Modern Humanity. London:

Cape

Sugiyama Y. 1981. Observations on the popu-

lation dynamics and behavior of wild chim-

panzees at Bossou, Guinea, 1979–1980.

Primates 22:432–44

Tomasello M. 1988. The role of joint atten-

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.

HUMAN ADAPTATION FOR CULTURE 529

tional process in early language develop-

ment. Lang. Sci. 10:69–88

Tomasello M. 1990. Cultural transmission in

the tool use and communicatory signaling

of chimpanzees? In “Language” and Intel-

ligence in Monkeys and Apes: Compara-

tive Developmental Perspectives, ed. S

Parker, K Gibson, pp. 274–311. Cam-

bridge, UK: Cambridge Univ. Press

Tomasello M. 1994. The question of chimpan-

zee culture. In Chimpanzee Cultures, ed. R

Wrangham, W McGrew, F de Waal, P

Heltne, pp. 301–17. Cambridge, MA: Har-

vard Univ. Press

Tomasello M. 1995. Joint attention as social

cognition. In Joint Attention: Its Origins

and Role in Development, ed. C Moore, P

Dunham, pp. 103–30. Hillsdale, NJ: Erl-

baum

Tomasello M. 1996a. Do apes ape? See Heyes

& Galef 1996, pp. 319–46

Tomasello M. 1996b. Chimpanzee social cog-

nition. Comment. Soc. Res. Child Dev.

Monogr. 61(3):161–73

Tomasello M. 1998. Reference: intending that

others jointly attend. Pragmat. Cogn.

6:219–34

Tomasello M. 1999. Perceiving intentions and

learning words in the second year of life.

See Bowerman & Levinson 1999. In press

Tomasello M, Barton M. 1994. Learning

words in non-ostensive contexts. Dev. Psy-

chol. 30:639–50

Tomasello M, Call J. 1994. Social cognition of

monkeys and apes. Yearb. Phys. Anthro-

pol. 37:273–305

Tomasello M, Call J. 1997. Primate Cogni-

tion. Oxford, UK: Oxford Univ. Press

Tomasello M, Call J, Nagell K, Olguin R, Car-

penter M. 1994. The learning and use of

gestural signals by young chimpanzees: a

trans-generational study. Primates 37:

137–54

Tomasello M, Call J, Warren J, Frost T, Car-

penter M, Nagell K. 1997. The ontogeny of

chimpanzee gestural signals: a comparison

across groups and generations. Evol. Com-

mun. 1:223–53

Tomasello M, George B, Kruger A, Farrar J,

Evans A. 1985. The development of ges-

tural communication in young chimpan-

zees. J. Hum. Evol. l4:175–86

Tomasello M, Gust D, Frost T. 1989. A longi-

tudinal investigation of gestural communi-

cation in young chimpanzees. Primates 30:

35–50

Tomasello M, Kruger A, Ratner H. 1993a.

Cultural learning. Target Artic. Behav.

Brain Sci. 16:495–552

Tomasello M, Savage-Rumbaugh S, Kruger

A. 1993b. Imitative learning of actions on

objects by children, chimpanzees and en-

culturated chimpanzees. Child Dev. 64:

1688–705

Tomasello M, Strosberg R, Akhtar N. 1996.

Eighteen-month-old children learn words

in non-ostensive contexts. J. Child Lang.

23:157–76

Tooby J, Cosmides L. 1989. Evolutionary psy-

chology and the generation of culture, Part

I. Ethol. Sociobiol. 10:29–49

Traugott E, Heine B. 1991. Approaches to

Grammaticalization, Vols. 1, 2. Amster-

dam: Benjamins

Visalberghi E, Fragaszy DM. 1990. Food-

washing behaviour in tufted capuchin mon-

keys, Cebus apella, and crabeating ma-

caques, Macaca fascicularis. Anim. Behav.

40:829–36

Vygotsky L. 1978. Mind in society. In The De-

velopment of Higher Psychological Proc-

esses. Cambridge, MA: Harvard Univ.

Press

Whiten A, Custance D. 1996. Studies of imita-

tion in chimpanzees and children. See

Heyes & Galef 1996, pp. 291–318

Whiten A, Custance DM, Gómez JC, Teixidor

P, Bard KA. 1996. Imitative learning of ar-

tificial fruit processing in children (Homo

sapiens) and chimpanzees (Pan troglo-

dytes). J. Comp. Psychol. 110:3–14

Whiten A, Ham R. 1992. On the nature and

evolution of imitation in the animal king-

dom: reappraisal of a century of research.

In Advances in the Study of Behavior, ed.

PJB Slater, JS Rosenblatt, C Beer, M Mil-

insky, pp. 239–83. New York: Academic

Ann

u. R

ev. A

nthr

opol

. 199

9.28

:509

-529

. Dow

nloa

ded

from

arj

ourn

als.

annu

alre

view

s.or

gby

Uni

vers

ity o

f B

ritis

h C

olum

bia

Lib

rary

on

08/1

9/09

. For

per

sona

l use

onl

y.